Skip to main content

Seminar in Linguistics (400-0-20)

Topic

Computational Methods for Social Understanding

Instructors

Robert Frederick Voigt Jr

Meeting Info

Annenberg Hall G01: Fri 1:00PM - 3:00PM

Overview of class

In this course we will explore modern methods for social understanding in computational linguistics. As contemporary NLP engineering progress has become dominated by uninterpretable neural models, a major question arising for computational sociolinguists and computational social scientists becomes whether and how these models can be used to generate insights for linguistic understanding of social phenomena. We will read, discuss, and build upon recent work addressing this challenging question with a focus on (and critical eye toward) methodological innovations leveraging embeddings, neural networks, and large language models.

Registration Requirements

Graduate standing or consent of instructor; ideally, coursework at the level of LING 334 or equivalent

Learning Objectives

Students in this course will gain knowledge on contemporary computational methods applied to social and linguistic questions, and a familiarity with contemporary work in this area. The course will have a strong focus on practical applications: students will engage with open-source code repositories and conclude the course with a project replicating or adapting these methods.

Teaching Method

Lecture, discussion, hackathons, and student led presentations

Evaluation Method

Self-evaluation; assignments include brief reading responses and a final project

Class Materials (Required)

No required texts (all will be provided by instructor)

Class Materials (Suggested)

N/A